Validating data-driven models at the network level

The aim of this thesis is to implement the data-driven models into a microscopic traffic simulator SUMO, using the observation data. As the observation data is already provided as for training, the main focus of the thesis will be based on the validating the data-driven model using the simulator software. To see the performance of the datadriven model, a reference model based on traditional approach is chosen and compared to two other flexible regression methods. These models should be applied to the chosen microscopic traffic simulator in a reasonable way. The data-driven models have to be examined, implemented and evaluated according to their optimum parameters for the given road network and observation data. In addition to that, the bias of the speed estimation, errors and performance measurements should be investigated.      «

The aim of this thesis is to implement the data-driven models into a microscopic traffic simulator SUMO, using the observation data. As the observation data is already provided as for training, the main focus of the thesis will be based on the validating the data-driven model using the simulator software. To see the performance of the datadriven model, a reference model based on traditional approach is chosen and compared to two other flexible regression methods. These models should be applied t...     »